Fine-scale land cover analysis of Luwire Wildlife Conservancy in Niassa Special Reserve, Mozambique for community-based conservation applications

Abstract

Niassa Special Reserve in Mozambique is one of the world’s largest contiguous Miombo woodland ecosystems, and home to megafaunal assemblages and human communities. Mozambique’s political history has impacted Niassa, with poaching, logging, and mining largely unchecked until the last decade. Niassa contains community-based natural resource management conservancies that fund local development through tourism and regulated hunting on land concessions, and thus support conservation. Forest cover loss totaling 4% between 2011 to 2020 was reported within the northern 3 provinces of Mozambique, including Niassa. How much forest loss has occurred within the Reserve itself, or conservation concessions specifically, is unknown. Regional land cover classification and forest loss data have been published at a 30m2 resolution, which obscures fine-scale changes and possible within-class modification, such as thinning, regrowth, or conversion. To fill these research gaps, we will conduct a 15-year land cover change analysis through image classification in Google Earth Engine, using 3.7m2 resolution Planet imagery centered on Luwire Wildlife Conservancy. Our analysis will show how much land cover change has occurred within Luwire, in other protected conservancy and non-conservancy lands, and outside the Reserve. This much finer spatial analysis will enable us to characterize within-class density by looking at individual tree crowns and understory cover. We will ground-truth these data in Luwire and explore the social forces impacting land cover. Our results will thus inform further research on land cover change and livelihood provisioning, ecosystem service payments to support local development, costs and benefits of farmland choices, and/or vegetation recovery pathways.